more precisely called branch of ai behind it

49
A HUMAN –LEVEL A HUMAN –LEVEL ARTIFICIAL INTELLIGENCE ARTIFICIAL INTELLIGENCE APPLICATION APPLICATION

Upload: allen-austin

Post on 12-Jan-2016

219 views

Category:

Documents


1 download

TRANSCRIPT

Page 1: More precisely called Branch of AI behind it

A HUMAN –LEVELA HUMAN –LEVELARTIFICIAL ARTIFICIAL

INTELLIGENCEINTELLIGENCEAPPLICATIONAPPLICATION

Page 2: More precisely called Branch of AI behind it
Page 3: More precisely called Branch of AI behind it
Page 4: More precisely called Branch of AI behind it
Page 5: More precisely called Branch of AI behind it

More precisely called

Branch of AI behind it

Page 6: More precisely called Branch of AI behind it
Page 7: More precisely called Branch of AI behind it

are Interactive games an area of Human-level AI research ?

is AI used in Interactive games ?

Page 8: More precisely called Branch of AI behind it

Picture Courtesy : Google Images

Page 9: More precisely called Branch of AI behind it

Human -like attributes expected Human -like attributes expected in a human-level ai system…in a human-level ai system…

Page 10: More precisely called Branch of AI behind it

are Interactive games an area of Human-level AI research ?

is AI used in Interactive games ?

Page 11: More precisely called Branch of AI behind it
Page 12: More precisely called Branch of AI behind it
Page 13: More precisely called Branch of AI behind it
Page 14: More precisely called Branch of AI behind it

Search

Planning

Logic

Page 15: More precisely called Branch of AI behind it
Page 16: More precisely called Branch of AI behind it

Focus : Game Tactics

A case study : the basicsA case study : the basics

How AI is used to enhance Game Tactics

How AI is used to enhance Game Tactics

AI tools used

Evolutionary computation &

Reinforcement Learning

Real-time Strategy Games

Page 17: More precisely called Branch of AI behind it

Genetic Algorithm

A learning technique with a mathematical reward function.

A learning technique with a mathematical reward function.

Page 18: More precisely called Branch of AI behind it
Page 19: More precisely called Branch of AI behind it

• Player needs to control armies to defeat all opposing forces in a virtual battlefield.

• Key to winning lies in efficiently collecting and managing resources., and appropriately allocating these resources over various action elements.

• Famous examples : Age Of Empires , World of Warcraft .Picture Courtesy : http://www.igniq.com/images/age_of_empires_3

Page 20: More precisely called Branch of AI behind it

Improve

Weaponry Attack

Page 21: More precisely called Branch of AI behind it

• AI in RTS games determines all decisions of the computer opponents.

• Encoded in the form of scripts. Called STATIC SCRIPTS

Page 22: More precisely called Branch of AI behind it
Page 23: More precisely called Branch of AI behind it
Page 24: More precisely called Branch of AI behind it
Page 25: More precisely called Branch of AI behind it
Page 26: More precisely called Branch of AI behind it

I don’t care about available

resources. Attack at earliest !!!Ha Ha Ha!!

I have to first well develop my army,

then only I can attack. This will

take a while.

HUMAN

AI

Picture Courtesy : World Of Warcraft

Page 27: More precisely called Branch of AI behind it

I have suffered heavy losses. Now I need to increase my

strength first. Small attacks are

of no use.

AI is gathering resources and preparing for

heavy assault.

HUMAN

AI

Picture Courtesy : World Of Warcraft

Page 28: More precisely called Branch of AI behind it

)()(

)(

1,,1,,

1,,

isisiaia

iaiai

SSSS

SSR

iaS ,

isS ,

winbSS

S

lostbSS

S

R

LsLa

La

LsLa

La

,max

,min

,,

,

,,

,

Page 29: More precisely called Branch of AI behind it

C end is a parameter and is set less than 0.5.

Contribution of State Reward is kept larger than Global Reward.

P max and R max are the maximum penalty and maximum reward respectively.

}{1)1(

1

}{)1(

max

max

bRb

bRC

b

bRCR

bRb

RbC

b

RbCP

Wi

endend

iendend

Page 30: More precisely called Branch of AI behind it

Evolutionary State Based Tactics Generator (ESTG)

Genetic Algorithm Application !!!

Counter Strategies are “played” against training scripts , only the fittest are allowed to the next generation.

Page 31: More precisely called Branch of AI behind it
Page 32: More precisely called Branch of AI behind it

Chromosome EncodingEA works with a population of chromosomes . Each represents a static strategy .

The chromosome is divided into the m states .

Start State 1 State 2 State m End

Page 33: More precisely called Branch of AI behind it

States include a state marker followed by the state number and a series of genes.

Chromosome Encoding

A Gene

Parameter values

Page 34: More precisely called Branch of AI behind it

4 types of genes

Page 35: More precisely called Branch of AI behind it

Partial example of a chromosome .

Chromosome Encoding

Page 36: More precisely called Branch of AI behind it

Fitness Function

b

MM

M

bMM

M

C

C

F

sa

a

sa

aT

,max

,minmax

Page 37: More precisely called Branch of AI behind it

Fitness Function

Page 38: More precisely called Branch of AI behind it

Genetic Operators

Page 39: More precisely called Branch of AI behind it

Genetic Operators

Page 40: More precisely called Branch of AI behind it

KT: State-based Knowledge Transfer

The possible tactics during a game mainly depend on the available units and technology, which in RTS games typically depend on the buildings that the player possesses.

Thus, we distinguish tactics using the Wargus states .

All genes grouped in an activated state (which includes at least one activated gene) in the chromosomes are considered to be a single tactic.

tactics

Page 41: More precisely called Branch of AI behind it

Extracting Tactics for a state

Page 42: More precisely called Branch of AI behind it

Performance of Dynamic Scripting Experiment Scenario

Page 43: More precisely called Branch of AI behind it
Page 44: More precisely called Branch of AI behind it

Performance Analysis

The three bars that reached 100 represent runs where no RTP was found (e.g., dynamic scripting was unable to statistically outperform the specified opponent).

The opponent strategies

Ave

rag

e R

TP

valu

eRTP is the number of the first game in which the adaptive agent outperforms the static agent.

low RTP value indicates good efficiency for dynamic scripting

Page 45: More precisely called Branch of AI behind it

Where we stand Where we stand today………today………

Achieved

Achieved

Achieved

Achieved

Achieved

Achieved

Not Achieved

Not Achieved

Page 46: More precisely called Branch of AI behind it

Picture Courtesy : Prince Of Persia , Google Images

Page 47: More precisely called Branch of AI behind it

DrawbacksDrawbacks

Giving undue advantages to AI agents.

Future – Scope:Future – Scope:

• Removing the “cheating” factor from Interactive games.

• Introduction of Creativity in AI agents.

• Capability of AI agents to reason with human-like Common Sense.

Page 48: More precisely called Branch of AI behind it

Ponsen,M. & Spronck,P.(2006). Automatically Generating Game Tactics via Evolutionary Learning.

Spronck,P. , Sprinkhuizen Kuyper,I. & Postma,E. (2004).Online adaptation of game opponent AI with dynamic scripting.

Sutton,R., & Barto,A.(1998). Reinforcement learning : an introduction.

Page 49: More precisely called Branch of AI behind it